STUDIES OF THE DEVELOPMENT OF PEPEROMIA HISPIDULA 39 1 
an independently developed one, if the forms possessing it were at all 
related genetically. It is a matter of fact, however, as has been pointed 
out before (Johnson, 1902, p. 336; Maneval, 1914, pp. 9, 10), that the 
forms possessing these composite sacs belong to widely separated fami- 
lies. Moreover this compound type of embryo sac may be found in 
but one or two genera of the family in which it occurs. Both these 
facts make it just as impossible to believe that the compound sac of 
Lilium, for example, is a long-established or fundamentally peculiar type 
as to believe that the very different peculiarities of the sac of Tulipa 
sylvestris, of Trillium grandiflorum, of Clintonia borealis, or of Smila- 
cina are primitive and fundamentally peculiar in their significance. 
The same reasoning would also deter us from regarding the abnor- 
malities of the sacs of Cypripedium (Pace, 1907), or Epipactis (Brown 
and Sharp, 191 1), as other than recently acquired variations of the 
usual type of development of the angiospermous embryo sac. 
Because of the scattered distribution of these abnormal types of 
embryo sac it seems quite clear that they have been independently de- 
veloped. Moreover, if the 4-spored, 8-nucleate type of Lilium can be 
developed phylogenetically from the i-spored, 8-nucleate type char- 
acteristic of its family as a whole, and therefore presumable primitive, 
then the 4-spored, i6-nucleate type of Peperomia may well have arisen 
from the 4-spored, 8-nucleate type of the other Fiperaceae, That is, 
the i6-nucleate sac has in all probability, contrary to the opinion of 
Ernst, arisen from the 8-nucleate one (see Johnson, 1900a, 1902). For 
the same reason the single synergid of Peperomia must be regarded as a 
later, incidental modification of the usual arrangement, of no more sig- 
nificance than the fact that Ornithogalum has but one synergid while 
the rest of the Liliaceae have two. 
Summary 
1. In vegetative structure Peperomia hispidula is the simplest 
described species of the genus and family. The delicate herbaceous 
stem, the single closed bundle in stem and root, and the delicate, 
gland-covered leaf, are probably not primitive features but are due to 
recent modification of the more complex type of structure, which is 
characteristic of the other members of the genus. 
2. This species has hermaphrodite, naked flowers and there is no 
indication in its development that it ever possessed floral envelopes. 
There are, as in other Peperomias, only 2 microsporangia per stamen 
